THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 275, No. 49, Issue of December 8, pp. 38554–38560, 2000 © 2000 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Expression of ␥- in and Its Interaction with the Smooth Muscle Sarcoglycan- Complex*

Received for publication, August 25, 2000, and in revised form, September 15, 2000 Published, JBC Papers in Press, September 18, 2000, DOI 10.1074/jbc.M007799200

Rita Barresi‡§, Steven A. Moore¶, Catherine A. Stolleʈ, Jerry R. Mendell**, and Kevin P. Campbell‡ ‡‡ From the ‡Howard Hughes Medical Institute, Department of Physiology and Biophysics and Department of Neurology, University of Iowa College of Medicine, the ¶Department of Pathology, University of Iowa College of Medicine, Iowa City, Iowa 52242, the ʈGenetic Diagnostic Laboratory, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104, and the **Department of Neurology, the Ohio State University, Columbus, Ohio 43210

The sarcoglycan complex in striated muscle is a het- phins (for reviews see Refs. 1–4). Recently, other such erotetrameric unit integrally associated with sarcospan as (5–7) and neuronal nitric-oxide synthase (8– in the -glycoprotein complex. The sarcogly- 10) have been shown to correlate with the DGC at the sarco- cans, ␣, ␤, ␥, and ␦, are mutually dependent with regard lemma. The interaction of the DGC with components of the to their localization at the , and (11–13) may have an important role in in any of the sarcoglycan lead to limb-girdle mus- force transmission and in sarcolemmal protection (14–16). The ␤ cular dystrophies type 2C–2F. In smooth muscle - and importance of the SGC in maintaining the sarcolemmal stabil- ␦ ⑀ - are associated with -sarcoglycan, a gly- ity is evident from the various forms of limb-girdle muscular ␣ coprotein homologous to -sarcoglycan. Here, we dem- dystrophy (LGMD) caused by mutations in any of the genes onstrate that ␥-sarcoglycan is also a component of the coding for the SGs (17–21). In addition, a functional role for the sarcoglycan complex in the smooth muscle. First, we SGC in signaling has been hypothesized (22, 23). show the presence of ␥-sarcoglycan in a number of Several mouse models of muscular dystrophy have been en- smooth muscle-containing organs, and we verify the ex- ␣ ␤ istence of identical transcripts in skeletal and smooth gineered in which -(Sgca-null) (24, 25), -(Sgcb-null) (26, 27), ␥ ␦ muscle. The specificity of the expression of ␥-sarcogly- -(Sgcg-null) (28), or -SG (Sgcd-null) (29, 30) genes have been can in smooth muscle was confirmed by analysis of disrupted. In addition to the pathology, all but smooth muscle cells in culture. Next, we provide evi- the Sgca-null mice show a cardiomyopathic phenotype. Fur- dence for the association of ␥-sarcoglycan with the sar- thermore, a spontaneous mutant, the ␦-SG-deficient BIO14.6 coglycan-sarcospan complex by biochemical analysis hamster, has been studied as an animal model for dilated and and comparison among animal models for muscular dys- hypertrophic cardiomyopathy (31, 32). Correspondingly, pa- trophy. Moreover, we find disruption of the sarcoglycan tients with defects in ␤-, ␥-, and ␦-SGs, but not ␣-SG, are complex in the of a patient with occasionally affected by dilated cardiomyopathy (33–35). ␥-sarcoglycanopathy. Taken together, our results prove In skeletal and , the SGC is a heterotet- that the sarcoglycan complex in vascular and visceral rameric unit composed of the transmembrane glycoproteins ␣-, smooth muscle consists of ⑀-, ␤-, ␥-, and ␦-sarcoglycans ␤-, ␥-, and ␦-SGs. The synthesis of all four of the proteins is and is associated with sarcospan. required in order to ensure the proper localization of the com- plex to the cell surface membrane (36); thus, the occurrence of a single mutated SG causes the loss or the reduction of the 1 The sarcoglycan-sarcospan complex (SGC) is part of the other SGs at the plasma membrane. dystrophin-glycoprotein complex (DGC), a group of proteins Whereas the expression of ␣-SG is restricted to striated well characterized in skeletal and cardiac muscle. The DGC muscle cells (37), a recently identified homologous , also includes dystrophin, (␣- and ␤-), and syntro- ⑀-sarcoglycan, is also expressed in several other tissues (38, 39). In skeletal muscle ⑀-SG interacts with ␤-, ␥-, and ␦-SGs consti- tuting a second SGC that co-exists with the conventional one * This work was supported in part by the Muscular Dystrophy Asso- (25), whereas in smooth muscle it associates with ␤- and ␦-SGs ciation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby and sarcospan (11, 40). Since most studies indicate the pres- marked “advertisement” in accordance with 18 U.S.C. Section 1734 ence of four SGs as a requirement for the functionality of the solely to indicate this fact. complex, it is expected that the smooth muscle SGC may either The nucleotide sequence reported in this paper has been submitted to the ␥ TM contain a yet unidentified sarcoglycan, a homologue of -SG, or DDBJ/GenBank /EBI Data Bank with accession number AF282901. ␥ § Supported by the Italian Telethon Grant 305/b and the Muscular -SG itself. Dystrophy Association. In previous studies the ␥-SG transcript and protein expres- ‡‡ Investigator of the Howard Hughes Medical Institute. To whom sion have been considered confined to striated muscle (18, 41). correspondence should be addressed: Howard Hughes Medical Insti- However, recent reports showed the presence of the ␥-SG tran- tute, University of Iowa College of Medicine, 400 EMRB, Iowa City, IA 52242-1101. Tel.: 319-335-7867; Fax: 319-335-6957; E-mail: kevin- script in lung (11). Nevertheless, attempts to immunolocalize [email protected]. ␥-SG in smooth muscle have failed, and the question whether 1 The abbreviations used are: SGC, sarcoglycan complex; DGC, dys- ␥-SG or a smooth muscle isoform existed was unanswered. trophin-glycoprotein complex; LGMD, limb-girdle muscular dystrophy; Addressing this issue, the present study demonstrates that ␣ mdx, dystrophin-deficient mice; SG, sarcoglycan; Sgca-null, -SG-defi- ␥-SG itself is a component of the DGC complex in smooth cient mice; Sgcd-null, ␦-SG-deficient mice; wt, wild-type; RT-PCR, re- verse transcriptase-polymerase chain reaction; RACE, rapid amplifica- muscle. By using immunoblot analysis of several smooth mus- tion of cDNA ends; PAGE, polyacrylamide gel electrophoresis. cle-containing organs from wild-type mice, a panel of antibod-

38554 This paper is available on line at http://www.jbc.org ␥-SG Expression in Smooth Muscle 38555 ies directed against different portions of ␥-SG is able to detect were run on an agarose gel; the band was excised, and the DNA was a protein of the same size as in the skeletal muscle (35 kDa). To cloned into pGEM®-T Easy vector (Promega) and sequenced using exclude the presence of an isoform of ␥-SG, we performed universal primers. 5Ј-RACE reactions were performed using the First ChoiceTM RLM- RT-PCR and sequenced two identical full-length cDNAs in RACE Kit (Ambion Inc.). A random-primed reverse transcription reac- skeletal and smooth muscle. Through 5Ј-RACE PCR, we were tion and nested PCR was then carried out to amplify the 5Ј end of the able to identify two alternatively spliced mRNAs in skeletal ␥-SG transcript. Two sense primers corresponding to the adapter se- muscle but only one form in smooth muscle. By analysis of quence were provided. Two antisense primers were used against nucle- cultures of human coronary artery smooth muscle cells, we otides 474–497 (5Ј-CAGGCGAAGTCCATCTGCTGTAAC-3Ј), and 380– Ј Ј demonstrated that not only the expression of ␥-SG actually 412 (5 -GCGAGATTCACAACGAGGATGGCGAGCAGGAGA-3 )ofthe mouse ␥-SG sequence. The PCR conditions were as follows: hot start; 35 occurs in smooth muscle but also the biosynthesis of the SGs in cycles, 1 min at 94 °C, 1 min at 60 °C, 1 min at 72 °C; 7 min at 72 °C in the smooth muscle is dependent on differentiation. In addition, the last cycle. The pool of the resulting PCR products was cloned and we demonstrated by sucrose gradient fractionation and co- sequenced as described above. immunoprecipitation of smooth muscle DGC that ␥-SG is part Smooth Culture—Human coronary artery smooth muscle of the DGC and, along with the other SGs, is absent in the cells, media, and supplements were purchased from Cascade Biologics, Inc. The cells were grown at 37 °C, 5% CO in Medium 231 supple- smooth muscle of Sgcd-null mice. Furthermore, we observed 2 mented with smooth muscle growth supplement. Proliferating cultures that the expression of the SGC was perturbed in the smooth at 70% of confluence were switched to differentiation medium (M231 muscle layer of the vasculature in a patient with a in and smooth muscle differentiation supplement) and processed for im- the ␥-SG . In summary, these results demonstrate that the munoblot as described above. SGC in smooth muscle cells is identical to the less abundant DGC Isolation and Sucrose Gradient Fractionation—Mouse skeletal SGC isoform of striated muscle fibers. muscle and lung tissues (5 g) were solubilized in 50 mM Tris-HCl, pH 7.4, 500 mM NaCl containing 1% digitonin in the presence of protease EXPERIMENTAL PROCEDURES inhibitors (24). The solubilized material was incubated overnight at 4 °C with pre-equilibrated wheat germ agglutinin-agarose beads (Vec- Animal Models—Wild-type C57BL/10J and mdx mice were obtained tor Laboratories). The beads were washed with 50 mM Tris-HCl, pH 7.4, from Jackson ImmunoResearch Laboratories, Inc. Colonies of Sgca- (24) 500 mM NaCl containing 0.1% digitonin and eluted with 0.3 M N- and Sgcd-null mice (29) were established and maintained at the Uni- acetylglucosamine in 50 mM Tris-HCl, 500 mM NaCl containing 0.1% versity of Iowa Animal Care Unit according to the animal care digitonin. The eluate was concentrated and loaded onto 5–30% sucrose guidelines. gradient and centrifuged with a Beckman VTi65.1 vertical rotor at ␣ Antibodies—Mouse monoclonal antibodies Ad1/20A6 against -sar- 200,000 ϫ g for2hat4°C.Fractions were collected and analyzed by ␤ ␤ coglycan, Sarc1/5B1 against -sarcoglycan, and 35DAG/21B5 against immunoblot as described (47). ␥ -sarcoglycan were generated in collaboration with Louise V. B. Ander- Co-immunoprecipitation Assay—100 ␮g of DGC isolated from skele- son (Newcastle General Hospital, Newcastle upon Tyne, UK). Mono- tal muscle and lung were diluted to 1 ml in immunoprecipitation buffer ␤ clonal antibody 43DAG/8D5 against -dystroglycan was generated by as described (25). Sepharose-bound protein G beads (Amersham Phar- ␣ L. V. B. Anderson. Polyclonal antibodies directed against -sarcoglycan macia Biotech) were equilibrated in immunoprecipitation buffer, and 50 ␤ ␥ (Rabbit 98), -sarcoglycan (Goat 26), -sarcoglycan (Rabbit 204, Rabbit ␮l were incubated for 30 min at 4 °C with each sample. The beads were ␦ 245, Rabbit 208, and Sheep 25), -sarcoglycan (Rabbit 229 and Rabbit then spun down, and the supernatant was incubated with the anti-␤-SG ⑀ 215), -sarcoglycan (Rabbit 232), sarcospan (Rabbit 217 and Rabbit affinity purified antibody (Goat 26) for4hat4°C.50␮l of Sepharose- ␤ 256), -dystroglycan (Rabbit 83), dystrophin (Rabbit 31), and utrophin bound protein G beads were added and incubated for an additional 2 h. (Rabbit 56) were described previously (24, 37, 40, 42–45). A commercial After four washes in lysis buffer, the beads were incubated for 10 min monoclonal antibody (clone 1A4, Sigma) was used to detect smooth at 65 °C in loading buffer. The bound proteins were then resolved by a ␣ muscle -. 10% SDS-PAGE and analyzed by immunoblot. Immunoblot Analysis of Total Homogenate—Tissue samples ob- Immunofluorescence Analysis—An 11-year-old male with clinical tained from wild-type, Sgca-null, Sgcd-null, and mdx mice were dis- signs and symptoms of LGMD was biopsied for diagnostic testing. Ϫ sected and snap-frozen in liquid nitrogen and stored at 80 °C until Samples from patients with myopathies unrelated to muscular dystro- use. Frozen tissues were processed as described (46). Briefly, the sam- phy were used as controls. Seven-␮m cryosections of these skeletal ples were solubilized in a SDS extraction buffer containing 80 mM muscle biopsies were analyzed by immunofluorescence as described ␤ Tris-HCl, pH 6.8, 10% SDS, 0.115 M sucrose, 1% -mercaptoethanol, 1 previously (24). Sections were observed under a Zeiss Axioplan fluores- mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, and 1 mM cence microscope (Carl Zeiss, Inc.). Digitized images were captured EDTA. The protein concentration in each sample was assessed with under identical conditions using PAX-it software (Midwest Information DCProtein Assay (Bio-Rad). Cell cultures were washed in phosphate- Systems, Inc.). buffered saline, pH 7.4, and directly solubilized in buffer. 150 ␮gof proteins were then analyzed by SDS-polyacrylamide gel electrophore- RESULTS sis. The gels were blotted onto nitrocellulose membranes (Immobilon- NC). Membranes were incubated with primary antibodies and detected Immunoblot Analysis Demonstrates the Presence of ␥-Sarco- by SuperSignal chemiluminescence (Pierce) after incubation with the glycan in Smooth Muscle—Previous reports have clearly dem- appropriate horseradish peroxidase-conjugated secondary antibody. onstrated the expression of ⑀-, ␤-, and ␦-SG in smooth muscle Ј RT-PCR and 5 -RACE PCR—Total RNA from control C57BL/10 skel- both by immunofluorescence and immunoblotting (27, 29, 40). etal muscle, heart, uterus, and aorta was extracted using RNAzol B (Tel-Test) according to the manufacturer’s instructions. For RT-PCR, 2 In contrast, because the immunostaining remains elusive, the ␮g of RNA from various tissues was reverse-transcribed with Moloney presence of ␥-SG in smooth muscle has been controversial. murine leukemia virus (Stratagene). We used three sets of overlapping In order to verify whether the ␥-SG in smooth muscle is the primers spanning the sequence of exons 2–8 of the mouse ␥-sarcoglycan same form as the striated muscle one, or a tissue-specific iso- TM cDNA (GenBank AB024922) as follows: nucleotides 277–297 (5Ј-CC- form, we analyzed several smooth muscle-containing organs Ј Ј ACGGTCACCGAGGGCACTC-3 ) and nucleotides 538–555 (5 -CACT- from wild-type mice. Total homogenate from uterus, lung, blad- CTGGAGCGTATTTC-3Ј); nucleotides 453–471 (5Ј-AATAGGAATGGG- TCACTTG-3Ј) and nucleotides 819–836 (5Ј-CTAAGGTCTTGAAATGG- der, and aorta, along with skeletal muscle from quadriceps G-3Ј); and nucleotides 762–780 (5Ј-GGCCAGAAGGAGCTCTTTT-3Ј) femoris, was tested by immunoblot for the presence of the and nucleotides 1240–1261 (5Ј-CCCCTGCATGCTTCTAAGTGTT-3Ј). 35-kDa ␥-SG band. A panel of five antibodies directed against Two sets of primers were used to amplify exons 1 and 2 as follows: different portions of the protein was used (Fig. 1a). All the A-sense, nucleotides 38–56 (5Ј-CCCTCATCGGCAATCAAGT-3Ј); A-a- samples were found positive, and the result was reproducible Ј Ј ntisense, nucleotides 381–398 (5 -CGAGGATGGCGAGCAGGA-3 ); with all the antibodies (Fig. 1b and data not shown). Although B-sense, nucleotides 91–108 (5Ј-CTATTGCTTCAGACTTTG-3Ј); B-ant- isense, nucleotides 423–440 (5Ј-ACATCACTTTCAGAATCC-3Ј). 40 the amount of protein loaded was the same for each lane, the PCR cycles were performed under the following conditions: 94 °C for 1 antibodies against ␥-SG consistently showed bands of different min, 60 °C for 1 min, and 72 °C for 1 min. The resulting PCR products intensity in the examined organs. This might be related to the 38556 ␥-SG Expression in Smooth Muscle

FIG.2. a, line a, proposed organization for the ␥-sarcoglycan gene. The structures of the transcripts are schematized respectively in lines b and c. Transcript b is expressed in skeletal and smooth muscle, whereas transcript c is weakly detected only in skeletal muscle with the set of primer B. b, partial sequence of the ␥-sarcoglycan mRNA. The putative sequences for exons 1a and 1b are highlighted in light gray and dark gray, respectively. The dashed line indicates primers A; the dotted line indicates primers B. The start codon is underlined and corre- sponds to the beginning of exon 2. These sequence data are available from GenBankTM/EMBL/DDBJ under accession numbers AB024922, NM_011892, and AF282901. FIG.1. a, a panel of five antibodies directed against ␥-sarcoglycan was used for immunoblot analysis. ␥ϪSG is a transmembrane protein with an extracellular carboxyl terminus and a single transmembrane such as the transmembrane domain and part of the extracel- domain (boxed). The mouse ␥ϪSG protein sequence is shaded on the lular domain have not been analyzed. A smooth muscle isoform regions recognized by our antibodies. The sequences corre- presenting variations in the amino acid composition might sponding to the fusion proteins against which Rabbit 208 and Rabbit 245 were produced are indicated by asterisks and underlined, respec- show a different structure and unique interactions with the tively. b, total homogenates from various mouse organs were electro- other components of the complex. Consequently, the epitopes phoresed on 10% SDS-polyacrylamide gel and transferred onto nitro- may become undetectable in situ but still detectable when the cellulose membrane. Identical membranes were probed with antibodies protein is unfolded in a gel under reducing conditions. against ␥-sarcoglycan 35DAG/21B5 (␥-SG) and ␤-dystroglycan Rabbit 83 (␤-DG). Variations in the size of the proteins are artifacts due to the In order to establish whether the protein sequences are the viscosity of the samples. same in skeletal and smooth muscle, we isolated total RNA from uterus, lung, bladder, aorta, and skeletal muscle from different composition of these organs in smooth muscle and wild-type mice. RT-PCR was performed using primers specific other tissues such as endothelium, epithelia, connective tis- for ␥-SG. The sequence of the PCR products was identical in sues, and blood. For example, after microdissection of aortic skeletal and smooth muscle samples. tissue to eliminate blood residue and the surrounding connec- The set of primers A and B was designed to detect the tive tissue, a thick wall of smooth muscle and a single thin presence of the two alternatively spliced exons 1 (GenBankTM layer of endothelial cells were isolated, and the sample ap- AB024922 and NM_011892). Whereas the PCR performed with peared to be more enriched in ␥-SG than the other non-skeletal the primers A amplified the expected product in skeletal and muscle samples. In contrast, ␥-SG seemed to be particularly smooth muscle samples, the reaction performed with the set of low in the lung, which contains a significant amount of epithe- primers B revealed a faint product only in skeletal muscle (not lium, vascular endothelium, and blood. An identical nitrocellu- shown). lose membrane was immunoblotted against ␤-dystroglycan, a In order to fully characterize the 5Ј end of the ␥-SG tran- protein present in skeletal and smooth muscle as well as in script, we performed 5Ј-RACE PCR on mouse skeletal muscle epithelial cells (Fig. 1b). In this case the intensity of the bands RNA samples. We successfully amplified a major product that was comparable in all samples but still relatively low in the corresponds to the exon 1 found by RT-PCR with primers A. lung. This suggests that the amount of blood in this organ None of the clones sequenced contained the sequence amplified contributes significantly to the total amount of protein present with primers B, indicating that this transcript is either less in the sample. stable or relatively rare in skeletal muscle. Our findings sug- For the following experiments we chose to use the mono- gest the presence of at least two exons 1 in the ␥-SG gene, as clonal antibody 35DAG/21B5. In view of the fact that ␥- and schematized in Fig. 2a. Exon 1a is mainly spliced with exon 2, ␦-SG proteins (both 35 kDa) share about 70% amino acid sim- generating a major transcript of 1119 base pairs, which is also ilarity (48), we performed peptide competition experiments, the only transcript present in smooth muscle. The genomic demonstrating that this antibody does not cross-react with analysis of this region was not extensively performed in the ␦-SG (not shown). present study. Since the protein coding sequence of ␥-SG starts Only One of the Two Alternatively Spliced ␥-Sarcoglycan from the second exon, and exon 1a and 1b correspond to an mRNAs Is Present in Smooth Muscle—All the antibodies we untranslated region, the alternative splicing of the mRNA is successfully used on immunoblot to detect ␥-SG failed in local- not expected to affect the final sequence of the protein. izing the protein by immunofluorescence. Although our five The Sarcoglycans Are Developmentally Expressed in Smooth antibodies cover the majority of the sequence of ␥-SG, regions Muscle—In view of the fact that the smooth muscle organs ␥-SG Expression in Smooth Muscle 38557

FIG.4.a, fractions 3–13 of the sucrose gradient separation of mouse lung DGC were resolved by 3–12% SDS-PAGE. After transfer the membranes were incubated with antibodies directed against ␤-sarco- glycan (␤-SG), ␥-sarcoglycan (␥-SG), ␦-sarcoglycan (␦-SG), ⑀-sarcogly- can (⑀-SG), sarcospan (SSPN), and ␤-dystroglycan (␤-DG). The peak of FIG.3.Lysates of human coronary artery smooth muscle cells the SG-sarcospan complex co-migrates in fractions 8–11 with the were electrophoresed and transferred to nitrocellulose mem- smooth muscle ␤-dystroglycan. b, DGC preparations from mouse lung branes. The membranes were probed with antibodies against the sar- and skeletal muscle were co-immunoprecipitated with an anti-␤-sarco- coglycans (␤-, ␥-, ␦-, and ⑀-SG), ␤-dystroglycan (␤-DG), and utrophin glycan antibody (Goat 26) and electrophoretically separated on 10% (UTR). A smooth muscle actin antibody (SMA) was used as a marker for SDS-polyacrylamide gels. Nitrocellulose transfers were probed with differentiation. Lane 1, undifferentiated cells. Lane 2, differentiated antibodies against ␣-sarcoglycan (␣-SG), ␥-sarcoglycan (␥-SG), ␦-sarco- cells at day 2. glycan (␦-SG), and ⑀-sarcoglycan (⑀-SG). In skeletal (Sk) and smooth muscle (Sm) ␥-SG co-immunoprecipitates with the other SGs. analyzed contain different cell types, we investigated the spec- ificity of the expression of the ⑀␤␥␦-SGC in smooth muscle by ␥-Sarcoglycan Is Absent in Smooth Muscle of Sgcd-null biochemical analysis of cultured human coronary artery Mice—It has been reported that the absence of ␤- and ␦-SGs smooth muscle cells. The cells were lysed and loaded on 10% leads to the loss of other components of the sarcoglycan-sarco- polyacrylamide gels at days 0, 2, 4, 6, and 8 after switch to span complex in smooth muscle (27, 29, 40). The disruption of differentiation medium. Fig. 3 represents the immunoblot the SGC has dramatic consequences for the smooth muscle of analysis of the samples at day 0 and 2, as no differences have the vessels, particularly in the heart where a pattern of con- been observed among the lysates after longer periods of differ- strictions of the coronary arteries is associated to ischemic-like entiation. Although ⑀-SG and ␤-dystroglycan have the same lesions finally leading to severe cardiomyopathy (27, 29). In strong level of expression before and after differentiation, the order to determine if ␥-SG is deficient along with the other SGs synthesis of ␤-, ␥-, and ␦-SGs dramatically increases in differ- and sarcospan in the vessels of cardiomyopathic mice, we an- entiating cells. alyzed aortas from wild-type, Sgca-null, Sgcd-null, and mdx ␥-Sarcoglycan Is an Integral Component of the SGC in mice by immunoblot for several components of the DGC and Smooth Muscle—In order to demonstrate that ␥-SG is part of a utrophin (Fig. 5). smooth muscle SGC that is similar to skeletal muscle, we As expected, the slight amount of dystrophin normally pres- performed sucrose gradient fractionation of smooth muscle ent in the aorta was absent in the sample from the mdx mouse, DGC isolated from mouse lung, the largest source of mouse whereas utrophin was strongly expressed in all the lanes. Also, smooth available in our laboratory. Proteins from predictably we did not observe ␣-SG in any of the samples in the sucrose gradient fractions were separated by electrophore- this study. The lack of ␣-SG expression demonstrated that sis on polyacrylamide gel. Immunoblotting with antibodies there was no contamination from striated muscle in the sam- against DGC components showed that the peak of the four ples from wild-type, Sgcd-null, and mdx mice. To microdissect smooth muscle SGs and sarcospan migrates in fractions 8–11 the aorta from the Sgca-null mouse, we used our standard (Fig. 4a). As none of the smooth muscle SGs, with the exception method. With this procedure we never detected ␣-SG in any of of ⑀-SG, is expressed in epithelial cells (11), we were able to the samples from other strains of mice. In the aortas from the analyze the smooth muscle SGC. This result confirmed that mdx mice, which occasionally show a cardiomyopathic pheno- ␥-SG is an integral part of the SGC in smooth muscle tissue. type at old age (49), the sarcoglycan-sarcospan complex was Furthermore, as already shown (11), ␤-dystroglycan migrates intact. In marked contrast, in the aortas of the Sgcd-null mice in fractions 5–7 as epithelial complex, and with the SG-sarco- all the SGs, including ␥-SG and sarcospan, were missing, span complex in fractions 8–11. whereas ␤-dystroglycan expression was unaffected. Similar re- The association of ␥-SG with smooth muscle SGC was fur- sults were obtained with the aortas from Sgcb-null mice (not ther demonstrated by co-immunoprecipitation of DGC isolated shown). from mouse lung using a polyclonal antibody against ␤-SG Sarcoglycans Are Similarly Reduced in Skeletal and Smooth (Goat 26). DGC isolated from mouse skeletal muscle was used Muscle of a LGMD-2C Patient—In order to characterize further as a positive control. The immunocomplexes were analyzed by the alterations of the SGC in smooth muscle, particularly in immunoblotting with antibodies directed against components regard to the human sarcoglycanopathies, we tested in parallel of the SGC as shown in Fig. 4b. In the smooth muscle ␥-SG skeletal muscle and vascular smooth muscle in the muscle co-immunoprecipitated with the other SGs. The absence of biopsies of several controls and one LGMD patient with a ␣-SG in the lung sample ruled out any contamination from mutation in the ␥-SG gene. skeletal muscle. Furthermore, in skeletal muscle the detection Genetic analysis of the LGMD patient revealed mutations of ⑀-SG immunoprecipitated with the anti-␤-SG antibody con- IVS2delϩ4-7 and T581C (L194S) in the ␥-SG gene. The IVS2 firmed the co-existence of the two distinct ␣␤␥␦ and ⑀␤␥␦ com- mutation changes the splice donor site from C(A/g)taagtatcat- plexes. These results indicate that the SGC complex in smooth tat to C(A/g)tatcattat. The effect of this mutation is unknown. muscle is constituted by ␤-, ␥-, ␦-, and ⑀-SGs. The skeletal muscle from the thigh had prominent, grouped 38558 ␥-SG Expression in Smooth Muscle

is generated by splicing of exon 1a. Interestingly, the genomic organization of ␦-SG, which shows 58% identity to ␥-SG at the DNA-level, displays three different untranslated exons 1 alter- natively spliced and differentially expressed in cardiac, skele- tal, and smooth muscle. A large deletion comprising exon 1b and 1c has been demonstrated to cause cardiomyopathy and muscular dystrophy in the BIO14.6 hamster (50). Although the differential expression of the transcript provides evidence for a tissue regulation in the ␥- and ␦-SG genes, the physiological significance of this phenomenon is still not clear. A major objection to the presence of ␥-SG in smooth muscle has been raised because of the unsuccessful immunolocaliza- tion. We speculate that in conjunction with ⑀-, ␤-, and ␦-SGs the antigens required for recognition by our antibodies in situ may be somehow masked, but they become more easily recognizable when the protein is unfolded on SDS-PAGE. The expression of the SGC in smooth muscle was explicitly demonstrated by biochemical analysis of smooth muscle cells in culture. As our interest was predominantly directed to vessel dysfunction leading to cardiomyopathy, we chose to analyze smooth muscle cells from human coronary arteries. Here we showed that all the components of the ⑀␤␥␦-SGC were ex- pressed. Furthermore, the synthesis of ␤-, ␥-, and ␦-SGs in smooth muscle cells dramatically increases with differentia-

FIG.5.Immunoblot analysis of aorta total homogenates. Sam- tion, as already described in skeletal muscle (40). ples from wild-type (wt), Sgca-null (␣Ϫ/Ϫ), Sgcd-null (␦Ϫ/Ϫ), and mdx Once we confirmed the presence of ␥-SG in smooth muscle, mice were analyzed by 10% SDS-PAGE and immunoblotted using an- our intention was to verify its interaction with the other DGC tibodies against the sarcoglycans (␣-, ␤-, ␥-, ␦-, and ⑀-SG), ␤-dystrogly- can (␤-DG), sarcospan (SSPN), dystrophin (Dys), and utrophin (UTR). components. To this end, we performed sucrose gradient frac- In the aortas of the Sgcd-null mice all the SGs and sarcospan are tionation and co-immunoprecipitation of DGC isolated from absent. To demonstrate equal loading of protein samples, we used a mouse lung, identifying ␥-SG in association with the other smooth muscle actin antibody (SMA). sarcoglycans and sarcospan. The animal models for limb-girdle muscular dystrophy types necrosis and regeneration, variation in fiber size from 10 to 100 2E and 2F lack expression of the sarcoglycans and sarcospan at ␮ m, and many internally placed nuclei. Endomysial fibrosis the smooth muscle plasma membrane, whereas the complex is and fatty replacement were mild. Immunofluorescence evalu- unaffected in LGMD-2D mice (27, 29, 40). In addition, perfu- ation of skeletal muscle revealed normal expression of dystro- sion of the vascular beds of the heart and diaphragm with ␥ ␣ phin; however, -SG staining was absent, and staining for -, Microfil, a liquid silicon rubber, showed focal vascular constric- ␤ ␦ -, -SG, and sarcospan was variably reduced (Fig. 6a). tions in Sgcb-null and Sgcd-null mice. Based on these data, a The smooth muscle of the vessels present in the same sec- novel hypothesis has been put forward that relates the car- tions displayed an equivalent pattern of perturbation of the diomyopathic phenotype to a dysfunction in the smooth muscle SGC (Fig. 6b); ␦-SG appeared to be the least reduced, whereas of the coronary arteries due to the disruption of the SGC in this traces of ␤-SG and sarcospan were detected. The weak fluores- tissue (29, 51). Mice deficient in ␥-SG also develop cardiomy- cence seen after incubation with the antibody against ␣-SG is opathy (28). Our finding that ␥-SG is a component of the considered specific since it is also visible on slides incubated smooth muscle SGC implies that vascular dysfunction may also only with secondary antibody. The analysis of ⑀-SG expression play a role in the pathogenesis of cardiomyopathy and muscu- was not performed because our antibody does not recognize the lar dystrophy in these mice. human protein on immunofluorescence. Further support for this hypothesis was obtained by compar- DISCUSSION ing the expression of the DGC in the aorta of three animal The objective of this study was to characterize the fourth models of muscular dystrophy. Only in the Sgcd-null aorta was component of the SGC in smooth muscle. Our results prove the SGC missing. Normal levels of SGC proteins in Sgca-null that ␥-SG is present in smooth muscle as well as in striated and mdx arteries are consistent with the mild or late onset of muscle and is associated with the previously described ⑀␤␦- cardiomyopathy displayed in these mice, respectively. To date, SGC. We also show that the loss of one of the other sarcogly- no reports have been published with regard to smooth muscle cans in smooth muscle affects the expression of ␥-SG. The expression of the other DGC components in the Sgcg-null reverse is also true; mutated ␥-SG protein is responsible for the mouse. Nevertheless, the occurrence of cardiomyopathy sug- lack of the remainder of the SGC. gests there may be disruption of the SGC in the smooth muscle Previous reports anticipated the presence of ␥-SG in smooth of this animal model. muscle (11, 27). Our data provide the final identification of this The vascular dysfunction may also have an important influ- protein as an integral component of the smooth muscle SGC. ence on the severity of muscular dystrophy in human patients. The fact that different antibodies recognized the protein on Here we showed that mutations in the ␥-SG gene destabilize immunoblot was not sufficient to exclude the existence of an the SGC in the vessels as well as in the skeletal muscle of an isoform of ␥-SG. However, the sequence of the RT-PCR prod- LGMD-2C patient, providing the first direct evidence of the ucts clearly revealed identical transcripts in skeletal and presence of the SGC in human smooth muscle tissue. The smooth muscle. Furthermore, whereas in skeletal muscle two disease severity in sarcoglycanopathies can vary from mild to transcripts have been identified, derived from the alternative severe. This variability has been correlated with the mutations splicing of exon 1a and 1b, the only transcript in smooth muscle in the sarcoglycan genes, e.g. missense mutations in both alle- ␥-SG Expression in Smooth Muscle 38559

FIG.6.a, cryosections of skeletal muscle biopsies from control and LGMD-2C patients were immunostained using antibodies against ␣-sarco- glycan (␣-SG), ␤-sarcoglycan (␤-SG), ␥-sarcoglycan (␥-SG), ␦-sarcoglycan (␦-SG), sarcospan (SSPN), and ␤-dystroglycan (␤-DG). ␥-Sarcoglycan is absent in the LGMD-2C patient, whereas the other sarcoglycans and sarcospan are variably reduced. Scale bar represents 20 ␮m. b, ␣-sarcoglycan (␣-SG), ␤-sarcoglycan (␤-SG), ␦-sarcoglycan (␦-SG), sarcospan (SSPN), ␤-dystroglycan (␤-DG), and smooth muscle actin (SMA) immunostaining in vascular smooth muscle of control and LGMD-2C patient. The same pattern of reduction of the SGs and sarcospan is shown in the vascular smooth muscle of the patient. In the vessels, the ␣-SG antibody staining corresponds to autofluorescent background. ␥-Sarcoglycan is not detectable by immunostaining in smooth muscle. Scale bar represents 20 ␮m. les are associated with a milder phenotype than homozygous other smooth muscle organs remains to be elucidated. Endo- null mutations (19, 20, 52). If the expression of the SGC is thelial cells release agents that could regulate the function of perturbed not only in the skeletal and cardiac fibers, but also in underlying vascular smooth muscle, thereby controlling vascu- the smooth muscle of the vasculature of these patients, vascu- lar tone by modulating the local concentration of vasoactive lar dysfunction might play a role in the pathogenesis of LGMD. substances, for instance synthesizing and releasing nitric ox- Therefore, the disruption of the SGC in the vessels is an ele- ide. The involvement of the ⑀␤␥␦-SGC in intercellular commu- ment that may be considered in the genotype-phenotype corre- nication is an intriguing hypothesis that may offer new ap- lation of the LGMD-2C/E/F patients. proaches for the treatment of the sarcoglycanopathies. Although the loss of the SGC leads to necrosis of the skeletal muscle, the smooth muscle organs of LGMD2C/F patients and Acknowledgments—We thank all the members of the Campbell lab- oratory for fruitful discussions, critical reading of the manuscript, and animal models show neither necrosis nor fibrosis. Moreover, in supply of reagents. We are indebted to Ramon Coral, Matt Hass, and the striated muscle of patients and mice with primary ␣-SG Steven Westra for expert technical assistance. We also thank Dr. James defects, there is no evidence of up-regulation in the expression Wilson for support of the genotype analysis. We are particularly in- of the ⑀␤␥␦-SGC, which might play a compensatory role. On debted to the patients and their families. All the DNA sequencing was carried out at the University of Iowa DNA core facility (NIH DK25295). account of these observations, it is tempting to hypothesize that Cell culture reagents were provided by the University of Iowa Diabetes the functions of the ␣␤␥␦-SG and ⑀␤␥␦-SG complexes may be and Endocrinology Research Center (supported by National Institutes divergent and to speculate that the ⑀␤␥␦-SGC may have a role of Health Grant DK25295). Initial biopsy screening and sarcoglycan in a still unidentified metabolic or signaling pathway rather mutational analysis in patients was supported by grants from the Muscular Dystrophy Association (LGMD Study to C .A. S., S. A. M., than mechanical transduction. In favor of this hypothesis, a and J. M.). number of mouse expressed sequence tag sequences similar to ␥-SG have been found in GenBankTM (accession numbers REFERENCES AV378726, AI121415, BB017993, and BB350604); the tissue 1. Henry, M. D., and Campbell, K. P. (1996) Curr. Opin. Cell Biol. 8, 625–631 2. Straub, V., and Campbell, K. P. (1997) Curr. Opin. Neurol. 10, 168–175 sources of the libraries were cecum, mammary gland, testis, 3. Ozawa, E., Noguchi, S., Mizuno, Y., Hagiwara, Y., Yoshida, M. (1998) Muscle and cerebellum. Interestingly, the transcript for ␥-SG, consid- & Nerve 21, 421–438 ered so far to be muscle-specific, seems to be expressed also in 4. Lim, L. E., and Campbell, K. P. (1998) Curr. Opin. Neurol. 11, 443–452 5. Blake, D. J., Nawrotzki, R., Peters, M. F., Froehner, S. C., and Davies, K. E. non-muscle organs, as has been shown with ⑀-, ␤-, and ␦-SG (1996) J. Biol. Chem. 271, 7802–7810 mRNAs (38, 39, 47, 53). 6. Sadoulet-Puccio, H. M., Khurana, T. S., Cohen, J. B., and Kunkel, L. M. (1996) Hum. Mol. Genet. 5, 489–496 The reason why abnormalities in smooth muscle SGC lead in 7. Grady, R. M., Grange, R. W., Lau, K. S., Maimone, M. M., Nichol, M. C., Stull, particular to vascular disturbance rather than dysfunction of J. T., and Sanes, J. R. (1999) Nat. Cell Biol. 1, 215–220 38560 ␥-SG Expression in Smooth Muscle

8. Brenman, J. E., Chao, D. S., Xia, H., Aldape, K., and Bredt, D. S. (1995) Cell 2535–2544 82, 743–752 31. Nigro, V., Okazaki, Y., Belsito, A., Piluso, G., Matsuda, Y., Politano, L., Nigro, 9. Brenman, J. E., Chao, D. S., Gee, S. H., McGee, A. W., Craven, S. E., G., Ventura, C., Abbondanza, C., Molinari, A. M., Acampora, D., Nishimura, Santillano, D. R., Wu, Z., Huang, F., Xia, H., Peters, M. F., Froehner, S. C., M., Hayashizaki, Y., and Puca, G. A. (1997) Hum. Mol. Genet. 6, 601–607 and Bredt, D. S. (1996) Cell 84, 757–767 32. Sakamoto, A., Ono, K., Abe, M., Jasmin, G., Eki, T., Murakami, Y., Masaki, T., 10. Chang, W. J., Iannaccone, S. T., Lau, K. S., Masters, B. S., McCabe, T. J., Toyo-Oka, T., and Hanaoka, F. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, McMillan, K., Padre, R. C., Spencer, M. J., Tidball, J. G., and Stull, J. T. 13873–13878 (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 9142–9147 33. van der Kooi, A. J., de Voogt, W. G., Barth, P. G., Busch, H. F., Jennekens, 11. Durbeej, M., and Campbell, K. P. (1999) J. Biol. Chem. 274, 26609–26616 F. G., Jongen, P. J., and de Visser, M. (1998) Heart 79, 73–77 12. Henry, M. D., and Campbell, K. P. (1999) Curr. Opin. Cell Biol. 11, 602–607 34. Moreira, E. S., Vainzof, M., Marie, S. K., Nigro, V., Zatz, M., and Passos- 13. Talts, J. F., Zeynepp, A., Goring, W., Brancaccio, A., and Timpl, R. (1999) Bueno, M. R. (1998) J. Med. Genet. 35, 951–953 EMBO J. 18, 863–870 35. Barresi, R., Di Blasi, C., Negri, T., Brugnoni, R., Vitali, A., Felisari, G., 14. Ervasti, J. M., and Campbell, K. P. (1993) J. Cell Biol. 122, 809–823 Salandi, A., Daniel, S., Cornelio, F., Morandi, L., and Mora, M. (2000) 15. Petrof, B. J., Shrager, J. B., Stedman, H. H., Kelly, A. M., and Sweeney, H. L. J. Med. Genet. 37, 102–107 (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 3710–3714 36. Holt, K. H., and Campbell, K. P. (1998) J. Biol. Chem. 273, 34667–34670 16. Weller, B., Karpati, G., and Carpenter, S. (1990) J. Neurol. Sci. 100, 9–13 37. Roberds, S. L., Anderson, R. D., Ibraghimov-Beskrovnaya, O., and Campbell, 17. Roberds, S. L., Leturcq, F., Allamand, V., Piccolo, F., Jeanpierre, M., Anderson, K. P. (1993) J. Biol. Chem. 268, 23739–23742 R. D., Lim, L. E., Lee, J. C., Tome´, F. M. S., Romero, N. B., Fardeau, M., 38. McNally, E. M., Ly, C. T., and Kunkel, L. M. (1998) FEBS Lett. 422, 27–32 Beckmann, J. S., Kaplan, J. C., and Campbell, K. P. (1994) Cell 78, 625–633 39. Ettinger, A. J., Feng, J., and Sanes, J. R. (1997) J. Biol. Chem. 272, 18. Noguchi, S., McNally, E. M., Ben Othmane, K., Hagiwara, Y., Mizuno, Y., 32534–32538 Yoshida, M., Yamamoto, H., Bo¨nnemann, C. G., Gussoni, E., Denton, P. H., 40. Straub, V., Ettinger, A. J., Durbeej, M., Venzke, D. P., Cutshall, S., Sanes, J., Kyriakides, T., Middleton, L., Hentati, F., Ben Hamida, M., Nonaka, I., and Campbell, K. P. (1999) J. Biol. Chem. 274, 27989–27996 Vance, J. M., Kunkel, L. M., and Ozawa, E. (1995) Science 270, 819–822 41. Yamamoto, H., Mizuno, Y., Hayashi, K., Nonaka, I., Yoshida, M., and Ozawa, 19. Lim, L. E., Duclos, F., Broux, O., Bourg, N., Sunada, Y., Allamand, V., Meyer, E. (1994) J. Biochem. (Tokyo) 115, 162–167 J., Richard, I., Moomaw, C., Slaughter, C., Tome´, F. M. S., Fardeau, M., 42. Holt, K. H., Lim, L. E., Straub, V., Venzke, D. P., Duclos, F., Anderson, R. D., Jackson, C. E., Beckmann, J. S., and Campbell, K. P. (1995) Nat. Genet. 11, Davidson, B. L., and Campbell, K. P. (1998) Mol. Cell 1, 841–848 257–265 43. Crosbie, R. H., Heighway, J., Venzke, D. P., Lee, J. C., and Campbell, K. P. 20. Bo¨nnemann, C. G., Modi, R., Noguchi, S., Mizuno, Y., Yoshida, M., Gussoni, E., (1997) J. Biol. Chem. 272, 31221–31224 McNally, E. M., Duggan, D. J., Angelini, C., Hoffman, E. P., Ozawa, E., and 44. Lebakken, C. S., Venzke, D. P., Hrstka, R. F., Consolino, C. M., Faulkner, J. A., Kunkel, L. M. (1995) Nat. Genet. 11, 266–273 Williamson, R. A., and Campbell, K. P. (2000) Mol. Cell. Biol. 20, 21. Nigro, V., de Sa Moreira, E., Piluso, G., Vainzof, M., Belsito, A., Politano, L., Puca, A. A., Passos-Bueno, M. R., and Zatz, M. (1996) Nat. Genet. 14, 1669–1677 195–198 45. Ohlendieck, K., Ervasti, J. M., Matsumura, K., Kahl, S. D., Leveille, C. J., and 22. Betto, R., Senter, L., Ceoldo, S., Tarricone, E., Biral, D., and Salviati, G. (1999) Campbell, K. P. (1991) Neuron 7, 499–508 J. Biol. Chem. 274, 7907–7912 46. Matsumura, K., Tome´, F. M. S., Collin, H., Azibi, K., Chaouch, M., Kaplan, 23. Crosbie, R. H., Lebakken, C. S., Holt, K. H., Venzke, D. P., Straub, V., Lee, J. C., Fardeau, M., and Campbell, K. P. (1992) Nature 359, 320–322 J. C., Grady, R. M., Chamberlain, J. S., Sanes, J. R., and Campbell, K. P. 47. Ervasti, J. M., Ohlendieck, K., Kahl, S. D., Gaver, M. G., and Campbell, K. P. (1999) J. Cell Biol. 145, 153–165 (1990) Nature 345, 315–319 24. Duclos, F., Straub, V., Moore, S. A., Venzke, D. P., Hrstka, R. F., Crosbie, R. H., 48. Jung, D., Duclos, F., Apostol, B., Straub, V., Lee, J. C., Allamand, V., Venzke, Durbeej, M., Lebakken, C. S., Ettinger, A. J., van der Meulen, J., Holt, D. P., Sunada, Y., Moomaw, C. R., Leveille, C. J., Slaughter, C. A., K. H., Lim, L. E., Sanes, J. R., Davidson, B. L., Faulkner, J. A., Williamson, Crawford, T. O., McPherson, J. D., and Campbell, K. P. (1996) J. Biol. R., and Campbell, K. P. (1998) J. Cell Biol. 142, 1461–1471 Chem. 271, 32321–32329 25. Liu, L. A., and Engvall, E. (1999) J. Biol. Chem. 53, 38171–38176 49. Lefaucheur, J. P., Pastoret, C., and Sebille, A. (1995) Anat. Rec. 242, 70–76 26. Araishi, K., Sasaoka, T., Imamura, M., Noguchi, S., Hama, H., Wakabayashi, 50. Sakamoto, A., Abe, M., and Masaki, T. (1999) FEBS Lett. 447, 124–128 E., Yoshida, M., Hori, T., and Ozawa, E. (1999) Hum. Mol. Genet. 8, 51. Gnecchi-Ruscone, T., Taylor, J., Mercuri, E., Paternostro, G., Pogue, R., 1589–1598 Bushby, K., Sewry, C., Muntoni, F., and Camici, P. G. (1999) Muscle & 27. Durbeej, M., Cohn, R. D., Hrstka, R. F., Moore, S. A., Allamand, V., Davidson, Nerve 22, 1549–1556 B. L., Williamson, R. A., and Campbell, K. P. (2000) Mol. Cell. 5, 141–151 52. Piccolo, F., Roberds, S. L., Jeanpierre, M., Leturcq, F., Azibi, K., Beldjord, C., 28. Hack, A. A., Ly, C. T., Jiang, F., Clendenin, C. J., Sigrist, K. S., Wollmann, Carrie, A., Recan, D., Chaouch, M., Reghis, A., El Kerch, F., Sefiani, A., R. L., and McNally, E. M. (1999) J. Cell Biol. 142, 1279–1287 Voit, T., Merlini, L., Collin, H., Eymard, B., Beckmann, J. S., Romero, N. B., 29. Coral-Vazquez, R., Cohn, R. D., Moore, S. A., Hill, J. A., Weiss, R. M., Tome´, F. M. S., Fardeau, M., Campbell, K. P., and Kaplan, J. C. (1995) Nat. Davisson, R. L., Straub, V., Barresi, R., Bansal, D., Hrstka, R. F., Genet. 10, 243–245 Williamson, R., and Campbell, K. P. (1999) Cell 98, 465–474 53. Nigro, V., Piluso, G., Belsito, A., Politano, L., Puca, A. A., Papparella, S., Rossi, 30. Hack, A. A., Lam, M., Cordier, L., Shoturma, D. I., Ly, C. T., Hadhazy, M. A., E., Viglietto, G., Esposito, M. G., Abbondanza, C., Medici, N., Molinari, Hadhazy, M. R., Sweeney, H. L., and McNally, E. M. (2000) J. Cell Sci. 113, A. M., Nigro, G., and Puca, G. A. (1996) Hum. Mol. Genet. 5, 1179–1186